Fuzzy-Tug of war structural active control for a seismically excited benchmark highway bridge

In this paper a new optimized active control algorithm based on combination of fuzzy neural network and a new highly efficient meta-heuristic population based optimization method extracted from Tug of War competition presents under different earthquake loads. The Efficiency of the proposed control method has been evaluated on the recently proposed nonlinear highway bridge benchmark, consist of nonlinear isolation bearings and nonlinear structural elements which equipped with the hydraulic actuators. A 5-layer neural network is used to obtain the control force. The neural network is utilized to approximate nonlinear rules of control. It gets instructions to the actuators installed between the deck and abutments. Stability of control laws to choose the parameters of the neural network are derived based on Lyapunov theory. The Results are presented in terms of a well-defined set of performance indices which is comparable to previous methods. The results show that the proposed controller method in spite of a simple description of the nonlinearities and non-detailed structural information can effectively reduce the responses of the bridge especially maximum of base shear, maximum of midspan displacement and maximum of acceleration. Also sensible decrease in responses such as maximum of ductility, dissipated energy and plasticity connections show that the proposed method is very effective in reducing structural damages.